Analyst June 1983 Vol. 108 pp. 685-690 685 Scanning Potential Stopped-rotation Voltammetry Joseph Wang” and Bassam A. Freiha Deflartment of Chemistry New Mexico State University Las Cruces NM 88003 USA The technique of scanning potential stopped-rotation voltammetry which is based on measuring the differences between currents with the electrode rotation switched on and off while the applied potential is scanned linearly, is described. Asymmetric rotation pulses without the achievement of the rotation “off” steady-state current are employed. The resulting modulated response is free of most background current components directly proportional to the analyte concentration and reproducible. Well defined current -potential graphs are obtained for ascorbic acid dopamine homovanillic acid and hexacyanoferrate(I1) ion at the micromolar concentration level.Extremely low background signals are achieved at a glassy-carbon disk electrode allowing a detection limit of 7 x 10-8 M of dopamine. The tech-nique is simple and suitable for automation. Keywords Scanning potential stopped-rotation voltammetry ; hydrodynamic modulation ; solid electrode ; anodic oxidation Increasing effort is being directed towards the development of sensitive and reliable voltam-metric techniques employing solid electrodes. Many important compounds that oxidise at potentials too positive for them to be determined at mercury electrodes are analysed at solid electrodes. Pulse voltammetry which effectively corrects for the charging current at mercury electrodes is not effective for trace analysis at solid electrodes because of the additional (un-corrected) background current component due to redox reactions of the surface.lS2 Hydro-dynamic modulation at forced-convective solid electrodes has been shown to be a feasible technique for obtaining voltammograms for low concentrations of electroactive compound^.^ The convection pulsing produces a pulsation only in the convection-dependent current and thus it is free from most background interferences.In this way the analytical information for trace compounds can be extracted directly from the modulated response. Most modulation approaches for batch electroanalysis have utilised the rotating disk electrode (RDE) because of its effective mass transport and rigorous theoretical treatment.A sinusoidally modulated RDE about a centre rotation speed value w, has been incorporated with continuous potential scans to obtain sensitive volt ammo gram^.^ The peak to peak amplitude of the change in rotation speed (Am) is usually a small fraction (up to 10%) of w,. The current transients produced by a square-wave step of the rotation speed of a disk electrode have been e ~ a l u a t e d . ~ ~ Similar steppings in the rotation speed have been incorporated (in pulsed-rotation voltammetrp) with pointwise changes in the applied potential. In previous work in this laboratory we advocated the utilisation of stopped-rotation modula-tion in which the rotation is switched on and off while maintaining a constant applied poten-tial.’ By measuring the resulting current differences at a number of discrete potentials, sensitive voltammograms were developed and plotted pointwise.The main advantages of this approach over sinusoidal or square-wave modulation are its inherent sensitivity i.e. lo0 yo modulation and simple operation (on - off switching vs. level changes). However it suffers from the relatively long time (around 20 min) required to record the complete voltammogram. This paper presents a technique termed scanning potential stopped-rotation voltammetry, which involves the imposition of a linear potential scan on a disk electrode the rotation of which is sequentially switched on and off. The current difference between the “on” and “off” positions is measured at fixed time intervals after switching occurs and plotted as a function of the potential.By incorporating stopped-rot at ion volt ammetry with continuous potential scans sensitive voltammograms are recorded within 3 min while maintaining the high sensitiv-ity of discrete potential stopped-rotation voltammetry. A relatively rapid non-steady-state modulation (frequency 2-4 s) is incorporated with scan rates of 2-5 mV s-l. Similar method-ology has been described recently for obtaining sensitive voltammograms in flowing streams, * To whom correspondence should be addressed 686 WANG AND FREIHA SCANNING POTENTIAL Analyst Vol. 108 utilising scanning potential stopped-flow voltammetry.8 As a stringent test of the method a glassy-carbon electrode is employed because of its relatively large background current .9J* The characteristics and advantages of scanning potential stopped-rotation voltammetry were elucidated in this study.Experimental Apparatus The 200-ml capacity electrochemical cell the rotating disk assembly and the measuring system (i.e. polarograph and recorder) have been described in detail earlier.? A 0.75-cm diameter glassy-carbon disk served as the working electrode in conjunction with a silver -silver chloride reference electrode and a graphite rod counter electrode. Reagents The chemicals and reagents used have been described in detail previously,ll except that millimolar stock solutions of homovanillic acid (Sigma Chemical Co.) were prepared fresh every day. The supporting electrolytes were 0.1 M phosphate buffer (pH 7.4) and 0.1 M potassium chloride solution. Procedure A 200-ml volume of the phosphate buffer was introduced into the cell and de-aerated for 8 min.The nitrogen delivery tube was then raised above the solution and pre-treatment of the working electrode was begun. This consisted of scanning the applied potential between + l . O and -1.0 V at 50 mV s-l for three cycles. Following this measurements were made on blank and analyte solutions. The electrode was held at a potential for the start of the scan (usually -0.2 V) and after 20-30 s an anodic linear potential scan was initiated. During the scan period stopped-rotation modulation was provided by manual on - off switching (every few seconds) of the rotation speed. Because of the slow current decay when the rotation is “off” (compared with the rapid achievement of the “on” current steady state) asymmetric rotation pulses [e.g.2 (on) and 4 (off) s] have been used for obtaining a large limited modu-lated response Ail while using short cycling periods. Results and Discussion The long response times of steady-state stopped-rotation voltammetry [about 3 s (rotation “on”) and 25 s (“off ”)?I preclude its incorporation with continuous potential scanning. Instead of waiting so long a non-steady-state technique can be applied to obtain large current amplitudes utilising short cycling times. It has been shown in our constant-potential stopped-rotation approach? that a non-steady-state modulation with 3 s “on” and 3 s “off” (i.e. 5-fold reduction in the period required for steady-state operation) results in a diminution in current of only 20%.The resulting non-steady-state current amplitude is linearly dependent on the analyte concentration. Similar compromise between sensitivity and speed has been utilised in stopped-flow8 and stopped-stirringl2 voltammetry. The feasibility of combining a non-steady-state stopped rotation modulation with linear potential scans is discussed below. Fig. 1 illustrates typical linear-scan stopped-rotation voltammograms (raw and filtered data) for the oxidation of 5 p~ ascorbic acid and 10 p~ dopamine. As the potential is scanned anodically increased current oscillations followed by plateau regions are observed. The filtered a.c. current - potential data show well defined waves with half-wave potentials of about +0.48 V (ascorbic acid) and +0.24 V (dopamine); the plateau regions start a t +0.68 and +0.44 V respectively.The ascorbic acid wave is spread over a wider potential region as expected from the irreversible nature of the oxidation. Despite its irreversible oxidation the ascorbic acid response is quantifiable in contrast to that obtained at the micromolar concentra-tion level with potential-pulse techniques. The corresponding background stopped-rotation voltammograms (not shown) show zero a.c. response as will be discussed later. The sensitivity of stopped-rotation voltammetry compares favourably with that obtained by other voltammetric techniques currently being used for trace analysis a t solid electrodes. In Fig. 2 stopped-rotation diff erential-pulse and linear-scan voltammograms for the oxidation of 7.5 x M hexacyanoferrate(I1) are compared.The hexacyanoferrate(I1) ion is often use June 1983 STOPPED-ROTATION VOLTAMMETRY 687 0.6 0.3 0 0.6 0.3 0 EN Fig. 1. Linear scan stopped-rotation voltammograms for (a) 10 p~ dopamine and (b) 6 p~ ascorbic acid. Phosphate buffer 0.1 M. Conditions were as follows : pH 7.4; scan rate 2 mV s-l; cycling times 2 (on) and 4 (off) s; and rotation speed (on) (a) 1600 and (b) 3600 rev min-l. The lower graphs are the responses as they come off the chart-paper, while the upper plots are the filtered responses (plotted point-wise by measuring the individual current oscillations). I I I I I I 0.8 0.4 0 ” 0.8 0.4 0 EN Fig. 2. Comparison of stopped-rotation voltammetry (A) diff erential-pulse voltammetry (B) and linear-scan voltammetry (C) with and (D) without electrode rotation.Conditions were as follows 75 PM hexacyanoferrate(I1) in 0.1 M potassium chloride; scan rate 5 mV s-l; rotation speeds (A C) 2500 and (B D) 0 rev min-1; cycling times (A), 2 (on) and 4 (off) s; and pulse amplitude (B) 50 mV. in the evaluation of new electroanalytical techniques. Among the techniques compared in Fig. 2 stopped-rotation voltammetry yields the most defined and quantifiable response. Owing to the very low ax. background current the analytical information can be extracted directly from the stopped-rotation voltammogram ; in diff erential-pulse or linear-scan voltam-metry background correction should be made. A response similar to B was obtained by rotating the electrode during the differential-pulse scan (not shown) The large differential-pulse background is due to changes in the redox state of surface functional groups that occur during the potential step.l,l3 The linear-scan background currents were composed of the double layer charging and the surface redox reactions.The stopped-rotation response is free of these background current components (which are non-convective). At the 1 x 1 0 - 5 ~ level only stopped-rotation voltammetry permits quantification (not shown) ; the differential-pulse and linear-scan analytical responses were obscured by the high background currents. The lower background current of differential-pulse voltammetry at carbon paste electrodes permits quantification of hexacyanoferrate(I1) a t the micromolar concentration 1e~el.l~ Similar comparison between the voltammetric techniques using 7.5 x M dopamine and a glassy-carbon electrode yielded graphs similar to those in Fig.2 (not shown). The scan rate used in the stopped-rotation measurements 5 mV s-l is similar to that commonly employed in differential-pulse voltammetry; thus a complete voltammogram is recorded within about 3 min. The current - time transient following the rotation stoppage results primarily from the relaxation of the diffusion layer thickne~s.~ The thickness of the diffusion layer when the rotation is on (assuming steady-state conditions) is given by Levich behaviour : = 1.61D”3~-1/2~1/6 . . . . - - (1) where D is the diffusion coefficient of the species w is the rotation speed (“on,’) and v is the kinematic viscosity.Theoretically under conditions of immediate rotation stoppage an 688 WANG AND FREIHA SCANNING POTENTIAL Analyst Vol. 108 absence of solution motion (i.e. conditions of linear diffusion) as the rotation is stopped the diffusion layer thickness increases with the square root of time. Thus a t any time t following rotation stoppage 8ofp is given by Combination of equations (1) and (2) with the general equation for the convective-diffusion limiting current (3) . nFADC 21 = - 8 where n F A D and C having their usual meanings leads to the following expressions for the difference in the currents at the “on” and “off” positions : In practice “Cottrell conditions” are not achieved because of the time lag associated with the deceleration of the motor4 and the resulting continuous solution motion.Quantitative evaluation is based on the linear correlation between the limiting modulated response and concentration expected from equation (4). Two separate experiments were per-formed to confirm this linearity. Stopped-rotation voltammograms obtained after successive concentration increments of dopamine are shown in Fig. 3. These four measurements are 0.7 0.4 0.1 EN Fig. 3. (B-E) stopped-rotation voltammograms ob-tained after increasing the dopamine concentration in 2.5-p~ steps along with (A) the corresponding background cur-rent. Conditions as in Fig. l ( ~ ) . X .,o I” k 0.2 pA I I I I 0.8 0.4 0 EN Fig. 4. Stopped-rotation voltammogram for (A) 1.0 p~ dopamine and (B) its supporting electrolyte (phosphate buffer pH 7.4) solution.Conditions as in Fig. 1 ( b ) June 1983 STOPPED-ROTATION VOLTAMMETRY 689 a part of six concentration increments from 2.5 to 15 p~ dopamine. The plot of limiting modulated response (measured a t +0.4 V) against concentration was linear with a slope of 0.316 pA p ~ - l (correlation coefficient 0.999 ; intercept -0.07 PA). The background voltam-mogram (A) is characterised by its very low modulated response indicating good correction for nonconvective currents and the absence of electroactive contaminants (the small and random current spikes are electrical noises due to rotation stepping). A similar calibration experiment for ascorbic acid [six increments from 5 to 30 p ~ . Conditions scan rate 5 mV s-1; frequency 3 s; rotation speed (on) and buffer as in Fig.l ( a ) ] yielded a linear plot (slope 0.265 pA p ~ - l ; correlation coefficient 0.997 ; intercept +0.14 PA). Fig. 4 illustrates the utility of stopped-rotation voltammetry for obtaining defined voltam-mograms for low concentrations of electroactive species. Examination of the data for 1.0 p~ dopamine (B) and its corresponding background current (A) shows that the limiting current amplitude is about 0.54 pA whereas the average noise level is 20 nA. A detection limit of about 70 nM of dopamine may be determined if it is based on a signal to noise ratio of 2. Other rotating disk assemblies may result with lower noise levels i.e. lower detection limits. A series of eight successive stopped-rotation voltammetric measurements of 10 p~ dopamine, carried out over a total time of about 1 h was used to evaluate the precision of the technique [conditions as in Fig.1 ( b ) ] . Reproducible voltammograms at the same potential regions, were obtained. The mean limiting current amplitude found was 5.65 PA with a range of 5.58-5.75 pA. The relative standard deviation calculated for this series 0.96% indicates the feasibility of obtaining reproducible current - potential data for low concentrations of electro-active species at solid electrodes. Fig. 5 demonstrates the potential of stopped-rotation voltammetry for trace analysis of mixtures ; stopped-rotation current - potential graphs (raw and filtered response) for a mixture of 7.5 p~ dopamine and homovanillic acid show two defined waves and plateau regions with 0.8 0.4 0 EN Fig.5. Stopped-rotation voltammo-gram for a mixture of 7.5 p~ dopamine and homovanillic acid in 0.1 M phosphate buffer (pH 7.4). Conditions as in Fig. l(b). The lower graph is the response as it comes off the chart-paper while the upper plot is the filtered response 690 WANG AND FREIHA apparent half-wave potentials of +0.22 V (dopamine) and +0.69 V (homovanillic acid). In general species with a 0.3-0.4 V difference in their half-wave potentials can be measured simultaneously (depending on the reversibility and number of electrons). Derivatised hydro-dynamic modulation approaches have been developed recently for improving the selectivity of multi-components ana1y~is.l~ The a.c. current is not affected by the non-convective-dependent d.c. background current a t relative extreme potentials ; a defined (combined) plateau is observed at potentials more positive than +1.0 V (where the oxidation of water occurs) indicating the feasibility of using the technique for obtaining voltammograms of depolarisers with high positive potentials.In conclusion scanning potential stopped-rotation voltammetry appears to be a valuable technique for trace electroanalysis of oxidisable compounds at solid electrodes. Similar data may be obtained in the cathodic region (provided that pure blank solutions are used) thus avoiding the inconvenience (and hazard) of the dropping-mercury electrode. Switching the rotation speed of the electrode on and off is experimentally the simplest way to modulate the RDE; therefore the technique can be easily placed under automated control and can be per-formed with any RDE assembly without the need for especially programming circuitry for superimposing sinusoidal or square-wave modulation.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References Kissinger P. T. Anal. Chem 1976 48 17R. Sokol W. F. and Evans D. H. A n d . Chem. 1981 53 578. Wang J. Talanta 1981 28 369. Miller B. Bellavance M. I. and Bruckenstein S. Anal. Chem. 1972 44 1983. Bruckenstein S. Bellavance M. I. and Miller B. J. Electrochem. Soc. 1973 120 1351. Blaedel W. J. and Engstrom R. C Anal. Chem. 1978 50 476. Wang J. Anal. Chem. 1981 53 1528. Wang J. and Dewald H. D. Anal. Chim. Ada 1982 136 77. Galus Z. and Adams R. N. J. Phys. Chem. 1963 67 866. Brumleve T. R. Osteryoung R. A. and Osteryoung J. Anal. Chem. 1982 54 782. Wang J. Anal. Chem. 1981 53 2280. Wang J. Anal. Chim. Acta 1981 129 253. Wang J. and Freiha B. A. Talanta in the press. Wang J. Talanta 1982 29 805. Received October 26th 1982 Accepted December 16th 198